Automatic speech recognizers (ASRs) and text to speech (TTS) synthesizers are speech resources commonly implemented in automated interactive systems, for example in automated voice response systems used in telephone call centers. ASRs and TTS synthesizers are resource consumptive, and oftentimes must be implemented across several machines that constitute a speech server solution. In order to reduce hardware costs, the speech resources are typically shared among many clients and allocated only when they are actually needed.
Media processing resource Control Protocol (MRCP) is a protocol that provides a mechanism for a client device requiring media processing resources, for example ASRs and TTS synthesizers, to control such resources on a network. In particular, MRCP defines requests, responses and events needed to control the media processing resources. Unfortunately, the MRCP standard exclusively relies on the client to distribute client requests. Indeed, MRCP lacks methods that can be implemented by a server to distribute the media processing resources across multiple nodes. Thus, even though a server is better suited for distributing media processing resources, distribution of such resources remains under client control.
It therefore would be beneficial to provide a technique which enables an MRCP server to distribute its own media processing resources.